My areas of interest lie broadly in Data Mining and Machine Learning. More specifically, I have worked in
- Privacy Preserving Data Mining
- Distributed Mining in Large-scale (P2P) Networks
- Web Mining and Social Network Analysis
-
Distributed data mining from large-scale distributed astronomy catalogsDeveloping distributed algorithms for solving astronomy problems using data distributed across multiple Virtual Observatories. Building a web-service for astronomers using Hadoop and Google sky.
-
Multi-objective optimization-based privacy model for distributed data miningModeling privacy requirements of different users in a distributed computation environment using a multi-objective optimization framework allowing personally tailored privacy specifications. Adapting differential privacy model for distribution-free threat modeling for the optimization.
-
Game-theoretic approach toward privacy preserving distributed data miningApplying game theory to analyze assumptions in existing privacy preserving data mining algorithms. Analyzing effect of collusion in secure multi-party protocols and designing mechanisms for achieving no-collusion equilibriums.
-
Distributed feature selection using ordinal statisticsDeveloping distributed algorithms for P2P networks using order statistics. Algorithm shows good accuracy with bounded message complexity
- Privacy Preserving Data Mining Algorithm Development Developing a new random encoding for classification which preserves privacy without compromising on the accuracy of the result
-
Bayesian Network Ensembles
Relevant Projects:
Develop efficient ways to combine bayesian networks. The final ensemble expected to be optimal and redundancy-free.
I did my M.S. thesis in Bioinformatic Visualization. Here is a copy of my thesis.